Repurposing cardiovascular disease prediction models for cancer.

INTRODUCTION

Population cancer screening detects the presence of early-stage disease rather than assessing future disease risk. We evaluated whether widely implemented cardiovascular disease (CVD) risk models can predict 10-year cancer risk, comparing them with the QCancer risk model.

MATERIALS AND

METHODS

We evaluated four CVD prediction models: QRISK3, the Pooled Cohort Equations (PCE), SCORE2 and SCORE2-OP. The models were recalibrated using 20% of the UK Biobank (UKB) cohort and tested in the remainder and in the Clinical Practice Research Datalink (CPRD). We gauged model performance by discrimination (c-statistics) and calibration (slope and intercept) and evaluated feature importance.

RESULTS

In the UKB test set, the c-statistics for incident CVD ranged from 0.71 to 0.74 (9,712 events). All CVD models achieved a c-statistic of 0.63 for any cancer (23,010 events) and showed CVD-equivalent discrimination for gastro-oesophageal, liver and biliary tree, laryngeal, renal tract, and lung cancers (c-statistic range: 0.70;0.81). The recalibrated CVD models showed near-perfect calibration (median intercept 0.01, Q1;Q3 -0.05;0.03 and slope 1.01, Q1;Q3 0.95;1.14). Performance in CPRD (393,622 cancer events) was similar: the median c-statistic, calibration intercept, and slope were 0.01 (95%CI 0.00;0.02), 0.09 (95%CI 0.03;0.20), and 0.04 (95%CI 0.01;0.14) higher, respectively, in CPRD than in UKB. After age, smoking status and systolic blood pressure were the most influential predictors of cancer risk.

DISCUSSION: Widely implemented CVD prediction models perform similarly to the less widely used QCancer models in the prediction of incident cancers. They may be used to inform cancer prevention and guide risk-stratified monitoring. The recalibrated models are available through an open source web application.